Modeling Human Behavior of Local Residents in the Aftermath of a Large Earthquake – Wide-Area Evacuation, Rescue and Firefighting in Densely Built-Up Wooden Residential Areas

Abstract

It is very important in disaster prevention planning to estimate the level of human damage after large earthquakes under various scenarios that takes into account the day of week, the time of the disaster, weather conditions, earthquake intensity, etc. There have been many previous studies based on the spatial characteristics of urban areas about evaluating protection against fires, evacuation risks, and the safety of evacuation routes to designated areas. However, no study so far has integrated models of property damage (building collapse, fire spread, and street blockage) and human behavior (rescue activities, firefighting activities, and wide-area evacuation behavior), and carries out simulations in order to analyze human damage in detail. In this paper, we present a survey of previous studies of the methods of evaluating urban-area characteristics, rescue and firefighting activities, and wide-area evacuation, all of which have been discussed as separate issues. We summarize the findings within the respective fields, their methods of evaluation and modeling, and identify their issues. Based on this survey, we point out that the construction of an integrated simulation model requires six important activities. They are to: 1) carry out evaluations on a microscopic scale at the block or street level; 2) use an evaluation index that allows a direct grasp of the expected level of human damage; 3) take into consideration many detailed and concrete disaster scenarios; 4) take into consideration the interactions among rescue participants, firefighting participants and wide-area evacuees, along with the effects of property damage; 5) incorporate the concept of time; and 6) set up comparative scenarios that allow the quantitative evaluation of the effects of various measures or policies. Therefore, it is necessary to construct a model based on the concept of multi-agent simulation (MAS).